The fee waterfall
Money leaves an account at several separate moments, and each moment belongs to a different party. The item price and the domestic postage to the warehouse belong to the marketplace seller. Any service or handling charge belongs to the platform. A photo set may be included or billed, depending on the platform and the account. Storage beyond a free window belongs to the warehouse. The international shipping line is billed when a parcel is submitted, and destination tax is collected at the far end. One waterfall, seven or eight taps, and only the first two are quoted anywhere near the product page. Nothing in the sequence is hidden, only spread out.
What makes the waterfall hard to compare is that only the first two taps are visible at the moment of buying. The item price and the domestic postage sit on the marketplace page, in front of the buyer, in the same place for every platform. Everything downstream appears later and somewhere else: charges on a platform account page, shipping quotes at parcel submission, storage notices when a clock runs down. A table that compares platforms on one line is comparing the only line that was never in dispute. Everything after the product page has to be found by asking the right page at the right time, and which page that is depends on the party doing the charging.
Shipping is the largest of the later taps because it is billed on measured weight and measured volume rather than on the number printed in a listing. Two parcels with identical item lists can bill differently if one is packed with filler and the other is not, and consolidation changes the measurement again by merging several shipments into one. That is the case for running a shipping weight estimate before submitting a parcel. It does not produce a carrier rate, but it puts a number on the part of the bill that the packing decision controls. The gap between an estimate and an invoice is worth reading before submission rather than after it, though few people look at it twice.
Free storage sits in a different category from every other tap, and that is the whole point of the title above. It is not a price. It is a deadline with a price attached to the far side of it. While the clock runs, the warehouse bills nothing, and a comparison table can print a clean number. When the clock stops, a charge appears that was never in the table, attached to items the buyer already paid for and cannot un-buy. The clock is also invisible on the product page, because it starts when an item reaches a warehouse rather than when it is bought. Deadlines are easy to ignore until the day they pass.
The ones you can remove
The removable charges are shaped like decisions, and the first of them is storage itself. Nothing about a warehouse clock is fixed at the moment of purchase. Shipping the parcel before the free window closes removes that charge entirely, and it is the only tap in the waterfall that a calendar can switch off. The awkward half of this is that deciding earlier costs money elsewhere, because the decision gets made before every item has arrived and been looked at. Shipping early is not free either, since a parcel submitted too soon leaves later arrivals to travel in a second parcel. That trade is the whole of the storage question, and it is a scheduling problem rather than a pricing one.
The second removable group is shaped like parcel construction. Consolidating four orders into one parcel bills one international line instead of four, and it also removes three separate handling steps from the sequence. Removing an item before submission is cheaper than shipping something unwanted and dealing with it later, though the removal step itself can carry a charge and the platform terms for it are not verified here. Splitting a parcel is the opposite move, and it is worth it only when a line limit forces the choice. Each of these moves happens inside the account, and none of them needs a negotiation with anybody. They are also reversible right up to the moment the parcel is submitted.
The third removable group is shaped like measurement. Declared weight and dimensions feed the shipping bill, and declared value feeds the duty calculation at the destination, so both are inputs a holder can influence with better information. The shipping weight estimate and the duty estimate exist for exactly this step. Neither one predicts a carrier invoice, but both reduce the chance of paying for a number nobody checked before submission. Better inputs do not lower a rate. They keep a wrong number from becoming the one that gets billed at the end. Neither tool knows what a warehouse scale will report, so they remove guesswork rather than risk, and a better input still beats a hopeful one on submission day.
What remains is not removable by any decision available to a buyer. The item price is set by the seller. Domestic postage inside the source country is set by the seller. Destination tax is set by the destination. Any platform service charge is set by the platform, and this ledger has verified that figure at no platform, so it stays unquantified here rather than estimated. The split is clean: price-shaped charges sit outside your control, and decision-shaped charges are the whole of what you control. A charge that cannot be named is not a small charge. It is an unknown, and unknowns deserve the same suspicion as large numbers. Naming the taps is the only real defence on offer here.
Before and after
Before an order exists, a reader is comparing published claims. Public comparisons describe one platform as offering a zero percent service fee and another as offering a long free storage window, with a hundred days quoted in more than one place. This ledger has not read those numbers on the platforms themselves, so it repeats them only as claims made elsewhere. Treat them as the shape of a marketing position rather than as a bill you can predict, because the terms that apply to a specific account are set per account. Read as a claim, a headline is useful. Read as a term, it is a guess. A claim is a starting point, not a contract.
After an order exists, the same reader is looking at something different: an account page, a photo set, a warehouse row, a countdown. This is the stage where fee questions get concrete, because every charge from the third tap onward appears against a specific item with a specific date beside it. The gap between the two stages is where most budget surprises live. A comparison read in the first stage describes a platform. A bill read in the second stage describes a parcel, and the two rarely match line for line. The account page is also the only place where a deadline carries a date rather than a policy, and a date is harder to market than a policy is.
The structural difference can be stated without inventing a single rate. An item that sits past a storage deadline picks up a charge that no comparison table can show, because the table has no idea how long the buyer will hesitate. Four small orders that become one consolidated parcel pay one international line instead of four. The same catalogue, ordered the same day and shipped in the same week, produces a smaller bill than the version that waits, and the difference comes from sequence rather than from platform choice. Nothing in that comparison requires a different platform, a different seller or a different catalogue. Only a different week.
That leaves a short working order for anyone rebuilding a sheet after a shutdown. Rebuild the addresses, confirm each one answers, and decide what to order within the free window rather than after it. Run the shipping weight estimate before submitting, and check the platform ledger page for which platforms are currently taking orders and on what date that status was last read. What this piece cannot give you is a price comparison, because no platform rate or storage term has been verified by this site. Structure travels. Prices do not. The ledger rows for Kakobuy and Sugargoo carry a check date and an operating status, and no verified fee figure. Anyone who hands over a finished cost table for an exact haul is estimating, whether or not the table admits it.